The Snap Sampler tm

Sampling VariabilitySampling Variabilityand Trend Monitoringand Trend Monitoring

North American Environmental Field

ConferenceTampa, Florida

January 15, 2007

Mr. Sandy Britt, PG, CHGProHydro, Inc.

Fairport, New YorkSandy.Britt@ProHydroInc.com

Fewer Variables with SNAP SAMPLER

30

35

40

45

50

55

60

65

70

0 4 8 12 16 20 24

Sampling events

Con

cent

ratio

n

Many Variables with Purge and Pour

30

35

40

45

50

55

60

65

70

0 4 8 12 16 20 24

Sampling events

Con

cent

ratio

n

Low noise data

High noise data

What are sources of data error and variability?

Laboratory?Hold timeCross-contaminationManual vs. autosampler

Collection Method?Bail purge and samplePump purge and sampleLow flow purge and samplePassive (noPassive (no--purge) samplingpurge) sampling

Sample Handling?Bottle fillingTransport

Dye source with gravity feed to injection port

Simulated 4-inch well

Upgradient/ influent reservoir

Piezometers

Dye port

Constant head reservoir

Homer’s water supply bucket

Influent supply from constant head reservoir Effluent drain

Effluent reservoir

DTSC Sand Tank Well Model“homogeneous” model well

Britt, SL, 2005, Testing the In-Well Horizontal Laminar Flow Assumption with a Sand Tank Well Model. Ground Water Monitoring and Remediation 25, no. 3 p.73-81

percent of initial dye concentration

33%

40%

35%

Flow weighted averaging

Flow-weighted averaging effect

∇ Water chemistry changes as a well is pumped

∇ Why does chemistry change?

∇ “Stagnant” water?

Or…

∇ A varying mix of water entering the pump?

Active (Purge) Sampling Methods

Varljen, et al., 2006, Numerical Simulations to Assess the Monitoring Zone Achieved during Low-Flow Purging and Sampling, GWMR, 26: p. 44-52

Analysis of

Steady-State

purging

Hydraulics controls flow, including water coming from beyond the screen zone

But purge time controls what water discharges from the pump

Traditional and Low Flow Purging Static vs. Dynamic

Stability parameters monitored…

Achieving “stability” depends on:

∇ Pre-purge well stratification∇ Stratification in the aquifer∇ Inflow characteristics∇ Shadow effects in the aquifer∇ Density effects

∇ Stability criteria∇ Barn door (e.g. 10%)∇ Reasonable measures (EC or

ORP +/- 10 units, without directional drift)

∇ How often you measure

Purge Equilibration

This is what “Stability”looks like

How long does purge equilibration take?

Achieving true “stability” (i.e. flow-weighted-average)depends on:

∇ Well diameter∇ Well length∇ Pump position∇ Contaminant position∇ Other heterogeneity

Quick Calculation:4” well, 10 foot screen, pump in the middle,

250 ml/min purge rate

(2.5L/ft)*5 ft/(0.25L/min) = 50 min (plug flow)

>50 minutes

Not too long????

What about Sampling Passively?

What is Passive Sampling?

Sampling without purgingSampling without purging>Deploy in advance>Deploy in advance>Relies on >Relies on ““passivepassive”” flushingflushing

Examples:Polyethylene Diffusion Bag sampler (USGS)Rigid Porous Pipe sampler (USGS)Dialysis Membrane Sampler (USGS)Gore Sorber (WL Gore)

Snap Sampler (ProHydro)

• Natural flow delivered to well

• Ambient (passive) mixing according to native flow dynamics

• Stratification testing?

• Sample from same position(the key to consistency)

Passive EquilibrationCan Limit Variables

The Snap Sampler is a dedicatedpassive sampling system

• Deploy double-ended bottles in an open position.

• Sample after short or long residence time in the well—1-2 weeks or 3 or 6 months

• Mechanical or electric trigger closes bottles in situ.

• Sample transfer is not requiredSample transfer is not requiredat the well head for VOCs-No exposure to air

How the Snap Sampler works….

Insert

40ml 125ml

Set Snap Cap

• Load & Set Snap Caps

How the Snap Sampler works…continued

Lower downhole

• Modular samplers allow up to 4 bottles per trigger

• Multiple triggers can be used for multiple sampling depths

Hang on Dock Ring Secure

• Mechanical or electric trigger

Attach Trigger

Seal in situ,

reduce surface handling

In Situ:

• Sample at the same position each sampling event

• Sample collection takes place submerged in the well

No well-head sample transfer required

• Very good correlations

• VOCS, 1,4-dioxane, anionsParsons, 2005, McClellan Air Force Base, Sacramento, CA

Example data

Spotting data trends may depend on consistency of your collection method…

Many variables for purge:Many variables for purge:actual concentration, pump position, pumping duration, stability criteria requirements, wind, temperature, site surface or ambient air contamination, bottle fill rate, pour technique, speed of bottle closure, filtration, transport, analytical variability.

Fewer variables Fewer variables with with in situ sealedin situ sealedsamples:samples: actual concentration, transport, analytical variability.

Interpretation of simple data trendssimple data trends is easier with less random error

Illustration only, not site data

Simple Declining Trend

20

25

30

35

40

45

50

55

60

65

70

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24

Sampling event

Con

cent

ratio

n

"Actual"

30% Random

2025303540455055606570

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24

Sampling event

Con

cent

ratio

n

10% Random

2025303540455055606570

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24

Sampling event

Con

cent

ratio

n

Interpretation of more complicatedmore complicated data trends is easier with less random error

Illustration only, not site data

Declining then Increasing Trend

2025303540455055606570

1 3 5 7 9 11 13 15 17 19 21 23

Sampling event

Con

cent

ratio

n"Actual"

10% Random

2025303540455055606570

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24

Sampling event

Con

cent

ratio

n

30% Random

2025303540455055606570

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24

Sampling event

Con

cent

ratio

n

Interpretation of very complicatedvery complicated data trends is easier with less random error

Illustration only, not site data

Complex Trend

2025303540455055606570

1 3 5 7 9 11 13 15 17 19 21 23

Sampling event

Con

cent

ratio

nActual

10% Random

2025303540455055606570

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24

Sampling event

Con

cent

ratio

n

30% Random

2025303540455055606570

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24

Sampling event

Con

cent

ratio

n

Example Data Variation

0

20

40

60

80

100

120

140

160

Sampling Quarter

TCE

mic

rogr

ams

per l

iter

purge 24.3 134 54.9 120 28.9 92.7 71 5.6 12.6

1 2 3 4 5 6 7 8 9

Example Data Variation

0

20

40

60

80

100

120

140

160

Sampling Quarter

TCE

mic

rogr

ams

per l

iter

Snap 56.9 72.6 64.1 103.0 39.7 65.6 40.9 50.8 50.4

1 2 3 4 5 6 7 8 9

TCE Example, site in Southern California

Range: 5.6 to 134Avg. RPD Q to Q: 100%Median RPD Q to Q: 94%

Private site, unpublished data

Range: 39.7 to 103Avg. RPD Q to Q: 36%Median RPD Q to Q: 35%

Purge Sealed in situ

Pump purge and bail sample Passive Snap Sample

TCE Example, Quarter to Quarter change

Example Data Variation

0

20

40

60

80

100

120

140

160

Sampling Quarter

TCE

mic

rogr

ams

per l

iter

Snap 56.9 72.6 64.1 103.0 39.7 65.6 40.9 50.8 50.4

purge 24.3 134 54.9 120 28.9 92.7 71 5.6 12.6

1 2 3 4 5 6 7 8 9

Relative Percent DifferenceQuarter to following quarter

0%

20%

40%

60%

80%

100%

120%

140%

160%

180%

1 2 3 4 5 6 7 8

Event

RPD

Snap Sample Pump Purge

TCE data superimposed Quarter to Quarter Relative Change

• Directional dynamic unchanged

• Quarterly concentration change less exaggerated

Overall statistics indicate differences in methods:

Purgen = 100 comparison pairs

Mean RPD Q to Q: 66%

Median RPD Q to Q: 51%

Mean % change Q-Q: 298%Median % change: Q-Q: 71%

Sealed in situn = 81 comparison pairs

Mean RPD Q to Q: 48%

Median RPD Q to Q: 37%

Mean % change Q-Q: 138%Median % change: Q-Q: 55%

Note: differences include the actual changes in

concentration…

• Reduced variation possible through consistent downhole passive sampling method

• In Situ sealed samples avoid error from surface handling

www.SnapSampler.com

Harbec Plastics Ontario, New York

“Technical Innovation with Environmental Responsibility”

• Passive method adds consistency by avoiding variables introduced during purge step

• Data trend more closely reflects downhole condition